Address PR #463 CI R1 (2 P3) — REPORTING parity + PDF reproducibility

CI codex Round 1 verdict was ✅ Looks good. Addressing the two actionable
P3 doc findings; the other 4 P3 findings were explicit "no fix required"
mitigation confirmations.

- P3 REPORTING.md parity: the diagonal-covariance-fallback Note still
  called the deviation "conservative" at REPORTING.md:324-337, but the
  new authoritative REGISTRY.md note (in this PR) describes it as a
  non-paper approximation that is NOT provably conservative (direction
  of the discrepancy with full Σ_22 depends on sign + magnitude of
  dropped correlations). Aligned the REPORTING note to match the
  REGISTRY language, kept the description of the BusinessReport
  well_powered → moderately_powered downgrade as a practical safeguard
  (not a proof of conservatism), and cross-referenced the REGISTRY
  authoritative deviation block.
- P3 paper-review reproducibility: the roth-2022-review.md header
  listed "PDF reviewed: papers/roth-2022.pdf" but the /papers/ directory
  is gitignored, so the path isn't reproducible from the repo alone.
  Replaced the path with the published-article citation + DOI + an
  explicit note that the PDF was reviewed externally and that the
  /papers/ working directory is gitignored. Added pointers to the
  jonathandroth.com author page for reproduction.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: igerber

docs/methodology/REPORTING.md

@@ -328,13 +328,20 @@ a library setting.
   `DiagnosticReport.pretrends_power` block records
   `covariance_source: "diag_fallback_available_full_vcov_unused"` in
   that case, and `BusinessReport` downgrades a `well_powered` tier to
-  `moderately_powered` before rendering prose. This is a known
-  conservative deviation from the documented "use the full pre-period
-  covariance" position — it prevents the diagonal approximation from
-  producing an overly optimistic "well-powered" claim when correlated
-  pre-period errors could tighten the MDV. The right long-term fix is
-  to teach `compute_pretrends_power()` to consume `event_study_vcov`
-  and `event_study_vcov_index`; until that lands this downgrade stays.
+  `moderately_powered` before rendering prose. This is a documented
+  deviation from the paper-derived "use the full pre-period covariance"
+  position. **Not provably conservative**: under Roth (2022)'s NIS
+  framework and the library's Wald form, the MDV/power objects depend
+  on the off-diagonals of Σ_22, and the direction of the discrepancy
+  between full-Σ_22 and diag(ses^2) depends on the sign and magnitude
+  of the dropped correlations — see the `**Note (deviation from paper
+  — diagonal pre-period VCV fallback):**` block under `## PreTrendsPower`
+  in `docs/methodology/REGISTRY.md`. The `well_powered → moderately_powered`
+  downgrade in BusinessReport reduces the chance of an overly optimistic
+  claim in practice, but it is not a proof of conservatism. The right
+  long-term fix is to teach `compute_pretrends_power()` to consume
+  `event_study_vcov` and `event_study_vcov_index`; until that lands the
+  downgrade stays.
 
 - **Note:** Unit-translation policy. BusinessReport does not
   arithmetically translate log-points to percents or level effects to

docs/methodology/papers/roth-2022-review.md

@@ -2,9 +2,9 @@
 
 **Authors:** Jonathan Roth
 **Citation:** Roth, J. (2022). Pretest with Caution: Event-Study Estimates after Testing for Parallel Trends. *American Economic Review: Insights*, 4(3), 305-322.
-**PDF reviewed:** papers/roth-2022.pdf (18 pages, content pages 1-15)
-**Review date:** 2026-05-16
 **DOI:** https://doi.org/10.1257/aeri.20210236
+**Source reviewed:** AER:I 4(3), 305-322 (18 pages, content pages 1-15). PDF was reviewed externally and is not committed to the repository (the `/papers/` working directory is gitignored). Reproduce by downloading the published article via the DOI above or from the author's page at https://www.jonathandroth.com/.
+**Review date:** 2026-05-16
 
 ---